Personal Finance, Accounting, and Tax Ideas

"How I see AI use on the user side is if they can’t be detailed in creating a help desk ticket good luck using a LLM. Luckily where I work we don’t heavily use AI yet. Claude and ChatGPT is used mainly by IT and finance. Not sure why finance requires it. I’m honestly scared to ask. For IT it has manly replaced googling and to quickly debug scripts/code. We also use image/video gen with Comfyui. Here is where AI has been time consuming and frustrating. Nodes and workflows are constantly not working/breaking and it can take hours to days of troubleshooting and applying fix after fix until it works. Then when the workflow does work. The output wasn’t up to standard and isn’t usable. Add in top constant updates and new model releases that break workflows that work just a few days before. IMO it’s a nightmare."

"For others, it might be one page that, you know, is simple and gets to the point. And I think, in general, just having a policy in place is more important than not having one. And so the things that should be included in that are what AI platforms that employees can use and under what licenses. So is that, you know, ChatGPT Enterprise? Is that Microsoft Copilot? Is that Cloud Enterprise? Making sure that they're authorized users under those licenses and that the company has vetted, you know, the platforms that the employees are accessing. It should also include what information can be inputted into the AI system, right? And what information should never be inputted into it, right? And so for some companies, that might mean no customer confidential information goes into AI, social security numbers, personal health and protected health information, trade secrets and sensitive financial information. Like that's just a list of things that you might consider having a hard ban on putting into AI as inputs. But, you know, it really just depends on the company's business, what AI platforms they're using and what those licenses say. You know, additionally in the policy, think about how you want AI and human interaction to exist in your company, right? Do you need to require human review on certain functions like HR, like legal, like customer communications, compliance? Like having a policy that says, like, you can't just use this AI platform to completely replace that human review may be necessary for certain companies. And then lastly, I think a good policy always includes some sort of audit procedure, right? To make sure that the policy is working, to make sure that employees are complying with the policy. And so figuring out, you know, on a regular basis, whether that's quarterly, annually, how do we audit this policy and make sure it still works for our business and that our employees are following it. And is there an off-the-shelf one that people can start with and tailor to their company's needs? That's a good question. I'm sure you could put it into chat GPT or cloud and have them create it for you."